Baseline and time-averaged fluid removal affect technique survival in peritoneal dialysis in a non-linear fashion

Abstract

The longevity of peritoneal dialysis (PD) is limited by technique failure and patient mortality. The authors assessed the influence of baseline and time-averaged fluid removal on patient, technique and death-censored technique survival. Peritoneal and total fluid removal was measured 1 month after commencing PD, then 6 monthly, in 225 incident patients (mean age 55.3+/-15.8 years, 52% male). A Cox proportional hazards model regression analysis was performed to identify variables independently predictive of technique and patient survival. Seventy (31.9%) patients were transferred to haemodialysis and 39 (17.63%) died. Technique survival was greatest in the middle tertile of baseline total fluid removal (mean survival time 3.5 vs 2.5 and 2.2 years for the lower and upper tertiles, respectively, log rank 6.5, P=0.039). The middle tertile of both baseline and time-averaged total fluid removal were significant predictors of PD survival (adjusted hazard ratio (HR) 0.476, 95% CI 0.286-0.795, P=0.005 relative to the upper tertile and HR 0.573, 95% CI 0.350-0.939, P=0.027 for baseline and time-averaged, respectively). Other significant variables on multivariate analysis were body mass index (HR 1.044 per kg/m2, 95% CI 1.005-1.084, P=0.028), creatinine (HR 0.999 per micromol, 95% CI 0.998-1.000, P=0.048) and residual Kt/V (HR 0.418, 95% CI 0.233-0.747, P=0.003). Patient survival was not affected by fluid removal. Patients with moderate total fluid removal both at baseline and throughout their PD career have improved technique survival. Attention should be paid to optimizing total fluid removal.